b6^

ROYAL IRISH ACADEMY TODD LECTURE SERIES

VOLUME XIL

EDWARD GWYNN

THE METRICAL DINDSHENCHAS

Pakt V.

DUBLIN

HODGES, FIGGIS, amp; CO.

LONDON: WILLIAMS amp; NOEGATE

1935

Price Ten Shillings and Sixpence.

EP

BY PONSONBY amp; GIBBS.

TO THE MEMORY OF

ROBERT ATKINSON

TABLE OF CONTENTS.

PREFACE

General Introduction :

I. The Manuscripts

II. The First Recension:

(a) nbsp;nbsp;nbsp;Book of Leinster

III. nbsp;nbsp;nbsp;The First Recension:

(b) nbsp;nbsp;nbsp;Bd.-Ed.

IV. nbsp;nbsp;nbsp;The Second Recension .

V. Grouping of the Second Recension;

(a) nbsp;nbsp;nbsp;Arrangement

VI. Grouping of the Second Recension :

(b) nbsp;nbsp;nbsp;The Texts

VII. L and the Reviser

VIII. 0’Duinn’s Poem

IX. Alphabetical Table

X. Origins of the Dindshenchas

Addenda Corrigenda

Index of Personal Names Index of Place Names .

Index of Collective Names Index of Pirst Linesnbsp;Glossary .

211

PREFACE

This volume closes an undertaking begun more years ago than I care to reckon. It would be tedious to offer excusesnbsp;for the long delays which have interrupted its progress. Inbsp;should like, however, to say a word in apology for one of itsnbsp;many shortcomings. A complete edition of the Dindshenehasnbsp;ought obviously to include both the prose and the metricalnbsp;versions, the more so that the prose is in part anterior tonbsp;the verse: the latter is, in fact, as is explained in my Introduction, to some extent founded directly upon the prose.nbsp;The reason for the procedure adopted in this edition is thatnbsp;when it was first undertaken the Prose Dindshenehas hadnbsp;just been published by Whitley Stokes in the Revue Celtique,nbsp;and it would have seemed presumptuous for a novice to offernbsp;a fresh edition of material which had recently been handlednbsp;by so eminent a scholar. There was nothing for it but tonbsp;assume that anyone who might wish to study the verse wouldnbsp;keep beside him Stokes’ edition of the prose. It is, however,nbsp;unfortunate that Stokes neglected almost entirely the oldernbsp;recension of the Dindshenehas, that of the Book of Leinster,nbsp;and also adopted a method which obscured the true order ofnbsp;the later recension, that of the Rennes codex and of mostnbsp;other manuscripts. It is to be hoped that some day anothernbsp;scholar will publish an edition of the prose and verse together;nbsp;if my attempt to prepare the ground is of some assistance innbsp;his task, I shall be well satisfied.

The Glossary appended to this volume makes no pretence to completeness. It is intended mainly as a help to studentsnbsp;pending the publication of the Royal Irish Academy’snbsp;Dictionary. Words which are sufficiently illustrated in the

PREFACE.

instalments of that Dictionary already published, or in Windisch’s Wörterbuch, or in Meyer’s Contributions to Irishnbsp;Lexicography, are as a rule omitted. At the same time, Inbsp;have thought it worth while to include examples of manynbsp;common words whose various uses seemed to require fullernbsp;illustration than those authorities supply. In marking thenbsp;genders of nouns, I have tried to follow the Middle Irishnbsp;usage, so far as this could be ascertained: the neuter, beingnbsp;in the twelfth century already obsolescent, is only marked innbsp;cases where it is actually found in my text. The Glossarynbsp;has also been used to call attention to rectifications of thenbsp;text and translation made in the long list of Corrigenda.

The Indexes of Personal and Place Names are, it is hoped, nearly exhaustive; in the identification of places I have hadnbsp;to rely as a rule on the works of O’Donovan, Reeves, andnbsp;Hogan, with occasional additions and corrections by recentnbsp;writers, and a few suggestions of my own. A full investigation of the whole available material is very badly needed,nbsp;but it could only be carried out by a combination of literarynbsp;research and field-work, organised on a large scale.

I had hoped to add to my Introduction a chapter on metres, but have never been able to find the time requirednbsp;for an adequate treatment of the subject.

I have inscribed on another page the name of Robert Atkinson, at whose suggestion, or rather command, this worknbsp;was first undertaken. He was in no way responsible for thenbsp;faults of plan and execution of which its author is painfully conscious. Such as it is, let it be an offering to thenbsp;memory of an inspiring teacher and a great scholar.

E. J. G.

Erratum. On p. 112, last line, for quot;lltli” read “12tli. ”

PART V.

GENERAL INTRODUCTION

TO

THE DINDSHENCHAS.

I.—THE MANUSCRIPTS.

§ 1. The Dindshenchas has been handed down in two recensions. Of the earlier, in its complete form in versenbsp;and prose, the sole surviving representative is preserved innbsp;the Book of Leinster (L), written about 1160. It occupiesnbsp;pp. 151-170 and 191-216 of the lithographed facsimile.

Parts of the prose sections of the same recension survive in two fragmentary MSS., both probably of the fifteenthnbsp;century, Rawlinson B. 506 in the Bodleian Library (Bd.)nbsp;and the Gaelic MS. xvi (also called Kilbride 12) in thenbsp;National Library of Scotland (Ed.). Besides the prose, Bd.nbsp;has also the poem Temmr V; Ed. has five poems, Temair V,nbsp;Loch Garman (part), Eithne (= Cam Furhaide), Tonnnbsp;Chlidm, and Tuag Inbir (part).

§ 2. The following 13 manuscripts contain more or less complete copies of the second recension, both prose andnbsp;verse:—

the Book of Ballymote (Royal Irish Academy 23. P. 12), pp. 349-410 of the facsimile,nbsp;the Rennes manuscript, ff. 90-125.nbsp;the Book of Ui Maine (R.Ir.Ac., Stowe D. II. Itnbsp;ff. 143-169.

the Book of Leean (R.Ir.Ac. 23. P. 2), pp. 461-525. the Yellow Book of Leean (Trin. Coll. Dublnbsp;H. 2. 16), pp. 438-455 of the facsimile.

R.Ir.Ac., Stowe D. 11. 2.

R.Ir.Ac., Stowe B, II. 2. A fragment.

R.Ir.Ac., Stowe B. III. 1.

Trin. Coll. Dubl. 1322 (H. 3. 3).

Trin. Coll. Dubl. 1295 (H. 2. 4), pp. 462-590. an 18th century copy of B.

Trin. Coll. Dubl. 1289 (H. 1. 15), pp. 409-532; another 18th century copy of B. ‘

B2

4 nbsp;nbsp;nbsp;GENERAL INTRODUCTION.

12 nbsp;nbsp;nbsp;= Trin. CoU. Dubl. 1317 (H. 2. 15 6), pp. 157-end,

a copy of H.

13 nbsp;nbsp;nbsp;V = R.Ir.Ac., Reeves 832, pp. 61-197.

Small groups of poems belonging in the main to the second recension occur in the following manuscripts:—

14 nbsp;nbsp;nbsp;S,. = R.Ir.Ac., Stowe D. IV. 2, ff. 55, 56.

15 nbsp;nbsp;nbsp;E Trin. Coll. Dubl. 1436 (E. 4. 1), pp. 85-88.

16 nbsp;nbsp;nbsp;H3 = Trin. Coll. Dubl. 1286 (H. 1. 12), pt. 2, pp. 97-

end; a copy of E.

17 nbsp;nbsp;nbsp;X = R.Ir.Ac., 23. N. 11, pp. 17-26.

A few detached copies of single poems occurring in other manuscripts have been mentioned in their places.

§ 3. It will be convenient to give here brief descriptions of the above manuscripts, ignoring 10-13, 16, and 17 asnbsp;unimportant, before entering on a discussion of the contentsnbsp;and arrangement of L.

B.—The Book of Baliymote was written about the year 1400: see Atkinson’s introduction to the facsimile fornbsp;particulars as to the date and the scribes. A leaf is missingnbsp;after p. 406 of the facsimile: the old foliation, probably duenbsp;to the original scribe, passes from 226 to 228.

Aspiration is marked by tiny dots, which are often invisible in the photographic reproduction, and in the MS.nbsp;itself are often hard to recognise with certainty. Longnbsp;vowels are regularly left unmarked; on the other hand thenbsp;vowel i is frequently distinguished by a diacritic stroke innbsp;order to prevent confusion between in and m, etc.

R.—This manuscript has been described by Dottin in Rev. Celt. XV. 79, seq. The Dindshenchas occupies ff. 90-125.nbsp;Stokes thought that this section “was probably written innbsp;the fourteenth or fifteenth century” (Rev. Celt. xv. 272),nbsp;and Meyer (Todd. Lect. xvii, introduction) is of the samenbsp;opinion. There is no scribal note that throws any lightnbsp;upon the date. The copy of the Dindshenchas is the worknbsp;of a single scribe, except the last page, as to which see V, § 2,

GENERAL INTRODUCTION. nbsp;nbsp;nbsp;5

infra. Two folios have been lost between f. 114 and f. 115 of the present numeration: (so Dottin, ut sup.; Stokes, innbsp;Rev. Celt. xv. 274, allows for the loss of only one folio, butnbsp;a calculation of the amount of material wanting and of thenbsp;number of lines per folio proves that Dottin is right).

The Royal Irish Academy possesses a .set of photographs of this copy.

M.—The copy in the Book of Ui Maine was written in 1394 by Paelan Mac Gabhann. It is fully described in annbsp;article in Eriu x. 68, seq., where I have .shown that thenbsp;21 articles following Druim Banrhrech are an interpolation,nbsp;mainly in prose, drawn from a source which did not belongnbsp;to the Second Recension, but was closely allied to Bd.-Ed.nbsp;Most of these articles are repeated by the scribe in theirnbsp;proper place according to the order of the Second Recension,nbsp;but a few {Bócmd, Sinann, Loch Garman, Cam Furbaide)nbsp;appear only in the interpolation (both prose and verse). Thenbsp;three articles following Ath Luain are in a different handnbsp;from the rest and of later date. The first of these, on Tuaimnbsp;Dd Gualann (Tuam), is not found elsewhere ; as much of thisnbsp;as is decipherable is printed in the Addenda to this volume,nbsp;p. 116, infra.

M has further a kind of appendix in which are mixed up with other matter three poems belonging to the Dindshenchas,nbsp;on Aenach TJchbad, Ailech III, and Loch Riach. The first ofnbsp;these does not appear in any other copy of the Dindshenchas:nbsp;it will be found in the Addenda, p. 115, infra.

The appendix ends with a copy of the poem by Gillananaem Ó Duinn which is printed in Eriu x. 76-89. I have sincenbsp;collated two other copies of this poem : see p. 81, infra.

The scribe of M is abominably careless. Of all texts of the Dindshenchas this has the lowest intrinsic value.

Lc.—The Book of Lecan was written, mainly in the years 1417-8, by three scribes, Gilla fsu Mac Firbhisigh, Murchadhnbsp;riabliach Ó Cuindlis, and Adam Ó Cuirnin. The copy of thenbsp;Dindshenchas is the work partly of Mac Firbhisigh (pp.nbsp;461-509), partly of Ó Cuirnin (i^p. 510-525). Certainnbsp;articles which are peculiar to this manuscript—Cam Fraich,

6 nbsp;nbsp;nbsp;GENERAL INTRODUCTION.

Ard na Riag, Inter Muada, Cam Amalgaid, Mag Tihra, Sliab Gam, Ess Riiaid II (poem only), and Loch Gile—arenbsp;all concerned with places in or near Co. Sligo, and not farnbsp;from Lackan on the river Moy, where most of the volumenbsp;was written. They are somewhat later in point of languagenbsp;than the main body of the Dindshenchas, and it seems probablenbsp;that they were composed by a native of Sligo, very likelynbsp;by one of the MacPirbis family. They cannot, however, benbsp;ascribed to either of the scribes of Lc., as the text which Lc.nbsp;presents abounds in corruptions and must therefore benbsp;derived from an earlier manuscript. Some local hand maynbsp;also be detected, perhaps, in the rearrangement of articlesnbsp;89 to 108 (see p. 42, infra), which cannot easily be accountednbsp;for by supposing an accidental disturbance of the order bynbsp;displacement of the leaves in Lc.’s model. The traditionalnbsp;arrangement may have seemed unsatisfactory to an editornbsp;or scribe who was intimately acquainted with this westernnbsp;region.

The manuscript is in excellent preservation and the handwriting of both the scribes of the Dindshenchas is firm and clear, except that in MacFirbis’ part it is often difficult tonbsp;distinguish m, in, and ni. The text is deplorable, beingnbsp;inferior even to that of Lc.’s near relative, S.

Y.—This copy constitutes one of the many disconnected fragments which together make up the (so-called) Yellow Booknbsp;of Lecan—a title which by right applies only to a smallnbsp;fraction of the whole collection. The copy of the Dindshenchasnbsp;is imperfect: more than half is wanting. What remains fillsnbsp;9 folios, now numbered as columns 401-436 (= pp. 438-455nbsp;of the photographic facsimile). The script seems to belongnbsp;to the fifteenth century, but there are no notes to give information as to the date or the scribe’s name. It is the worknbsp;of one hand, excepting the article on Sliab Bladvia and thenbsp;verse (not the prose) of Ceilbe, which are added in a laternbsp;hand, on two strips of vellum. The text of the Dindshenchasnbsp;has been corrected throughout by a late hand—perhaps Chai’lesnbsp;O’Conor, who has similarly maltreated the section of thenbsp;MS. containing coll. 573-958. His corrections sometimesnbsp;obliterate the original reading.

GENERAL INTRODUCTION.

Y has lost its first folio and now begins at § 26 of the prose introduction to Temair : see Rev. Celt. xv. 282. Thenbsp;lost leaf probably contained (1) the prose paragraph Senchasnbsp;Dinid Érenn: (2) verse, Temair V : (3) prose, “Temair dono,”nbsp;ol Anuiirgen fri Fintan, etc.; (4) verse, Temair I: (5) prose,nbsp;“Temair dono” ol Armairgen A. mur Tea, etc., = Rev. Celt.nbsp;XV. 277, §l-§4; (6) verse, Temair II; (7) prose, Dindgnainbsp;Temrach: NemTiach, etc., = Rev. Celt. xv. 280, § 5-§ 25.nbsp;This is the order in SS2S3H, which differs from the arrangement in BR (Rev. Celt. xv. 277, seq.) and from that in Mnbsp;(Erin x. 71-2).

S.—Stowe D. II. 2, vellum, ff. 90 (the number 33 is repeated, so that the subsequent numbers are too low by 1). One leafnbsp;has been lost after f. 7, one after f. 33, one after f. 39 [38],nbsp;and one after f. 44 [43]. The volume is the work of a singlenbsp;scribe, Muiris 0 Clérig, who gives his name on f. 88 [87] v°.nbsp;(see part iv. 298). He may be Muiris mac an Ghiolla riabh-aigh ua Cléirigh, saai hi senchas 7 hi Uighionn, who died innbsp;1573 (P.M. V., p. 1674). Otherwise, the date of the manuscript is not easy to determine: the script is calligraphic,nbsp;ornamented with coloured initials, no doubt imitated from somenbsp;older exemplar; but the spelling points to the sixteenth ornbsp;seventeenth century, e.g., hhfuil, caroid (= carait), faighisnbsp;{= foods), flegh {= fledh), mogh (= mod), ghénamh {= dhén-anih). The volume belonged formerly to Charles O’Conor ofnbsp;Belanagare, who has written some notes in the margins andnbsp;has made on the last page an entry recording the death ofnbsp;his father, Donnchadh mac Cathail óig, on January 30“ 1750,nbsp;in his 76“ year, and his burial at Ballintubber, and givingnbsp;his pedigree up to Cathal croh-derg 0 Conor. The distinctivenbsp;feature of this MS. is the series of supplementary articlesnbsp;which are printed (with certain omissions, duly noted) innbsp;part iv. 268-310.

Sg.—Stowe B. II. 2, vellum, ff. 8. A fragmentary copy, containing only the beginning of the Dindshenchas: writtennbsp;perhaps in the fourteenth, certainly not later than thenbsp;fifteenth century. The recto of f. 1 is almost entirely illegible,nbsp;the verso of f. 8 completely so. Scribe unknown.

GENERAL INTRODUCTION.

S3.—Stowe B. III. 1, paper, ff. 104. The date is given by a note at foot of f. 89r°: Ag sin a MJieic Con i Dhomnaillnbsp;na tri line do iarrais oram dto scriohhadh duit. A mBailenbsp;Meic Cathail damh occa sccriobhadli an 28. Decimbir M.dc.l.iu.nbsp;1654. This note is in a beautifully neat script, beyond doubtnbsp;the same as that of the Academy’s copy of 0’Clery’s Leabharnbsp;Gabhdla, classed 23 K 32, and also of 23 D 17. Accordingnbsp;to O’Curry (MS. Catalogue, first series, vol. i, p. 84) it isnbsp;the hand of Cu Coigriche 0’Clery, whose autograph willnbsp;(dated 1664) is contained in 23 D 17, p. 271.

0 ’Clery wrote about a third of the volume: ff. 2v°, 15r° ad calc, to 18v°, 22v° to 46, 87r° (part) to 89v°, 102v° tonbsp;103r° (ds. of Léige); and also added the saints ’ names innbsp;margin of f. 104. The rest of the manuscript is in the handnbsp;of a pupil who tries to imitate his master. 0’Clery hasnbsp;corrected his work here and there.

E.—Trinity College, Dublin, no. 1436 (E. 4. 1), pp. 85-88, vellum. These are two leaves bound by chance into a medicalnbsp;manuscript. The handwriting and spelling indicate thenbsp;fourteenth or perhaps the fifteenth century. There are tennbsp;articles, both prose and verse: Sliab Bladma (acephalous).nbsp;Fid nGaiblc, Mag Liphe, Berba, Móin Gai Glais, Fafann,nbsp;Almu II, Alend, Carmun, Bóand II (four st., incomplete).nbsp;These are in the regular order, except for the inversion ofnbsp;Bóand I and II, which is also found in S. The two leavesnbsp;are therefore a fragment of a complete copy of thenbsp;Dindshcnehas.

The relation of E to other copies is peculiar. Its most remarkable feature is the inclusion of a long passage in thenbsp;poem on Carmun, which is also in L, but not in any othernbsp;member of the Second Recension. At other points also innbsp;this poem E agrees with L against all other copies: seenbsp;critical notes on lines 4, 9, 19, 21-24, 37, 39, 45, 51, 53-56,nbsp;58, 59, 68, 84, 85, 90; and compare further the readings atnbsp;81. Bladma, 4, 8, 17-20, and Mag Life, 2.

On the other hand, E has close relations with the S-grouj) of manuscripts, particularly with S: see VI § 3. As alreadynbsp;remarked, B agrees with S in placing Bóand II before Bóand I.

A comparison of the prose versions in E with those of L,

GENERAL INTRODUCTION.

Bd.-Ed., and R (representing the Second Recension) yields the following results :—

Sliab Bladma.—The part that remains agrees with Bd.-Ed. Fid nGaihU.—Nearest to L, but agrees in one point with Bd.-Ed. against LR, and in two points with R againstnbsp;L Bd.-Ed.

3Iag Life.—^Nearest to R, but omits R’s last paragraph, as do L Bd.-Ed.

Berha.—Nearest to R, with omission of one line, omitted also by L Bd.-Ed.

Main Gai Glens, Fajfann, Alnm II, Alend.—Agrees with R.

Of these, only Almu II is in L, none in Bd.-Ed. Carnmn.—-Agrees with R, except at end, where E has threenbsp;lines which are in L but not in R (following innbsp;Herinn foraih)-, E also agrees with L in adding danbsp;feraib after mochUitJie, and Étgied écruthach 7 maüenbsp;after rig ócca.

H.—Trin. Coll. H. 3. 3 (1322), ff. 72, vellum. Written mainly by Sean Ó Cianain, at Ard Choill: see note atnbsp;f. Ilby: Misi Sean 0 Cianain dosgriobh in dara la ria Notlaic,nbsp;et is ole Hum a tliinnim ar medugud ag Sean 0 Maolchonaire,nbsp;San Ard Coill damh. There are similar notes at 5az, 17flz,nbsp;29ÖZ, 36?)z, 375z, 38bz. The illness of Sean 0 Maolchonairenbsp;is mentioned again at GObz. O’Donovan in his (incompletenbsp;and unpublished) description of the Trinity College MSS. saysnbsp;that this 0 Mulconry “kept a celebrated school at Ardkyle,nbsp;near Sixmilebridge in Co. Clare, in the reign of Elizabeth.”nbsp;Sean 0 Cianain also wrote Rawlinson B. 506, see ZCP xii.nbsp;358. At 30bz there is a note Misi Tadg 0 Cietiain, presumably by a brother or kinsman of Sean 0 Cianain, and thisnbsp;man may have written part of the MS.; but, if so, his handnbsp;is scarcely distinguishable from that of Sean.

Other scribal notes mention the names Cu Chonnacht (46bz), Corbmac (58az), Sean Mac Conmara (53aw), Domnallnbsp;Mac Conmara (58amp;z), Tadg 0 Duuthach, i.e. 0 Dubhthaighnbsp;(lOamp;z).

This is the “vellum written about the year 1560 for John 0’Mulconry” mentioned by O’Donovan in the Ordnance

GENERAL INTRODUCTION.

Survey of Co. Londonderry, p. 223, whicli Stokes (Rev. Celt. XV. 274, note) supposed to be separate MS. Stokes wronglynbsp;ascribed H to the fifteenth century.

The scribe has certain peculiarities of spelling, which must be taken into account in judging of variant readings.

He frequently writes e for a, as in ier, dieta, for iar, diaid, etc.; 0 for a, as in catJio, athor, luon; oa, oai, for ua, uai,nbsp;as toatli, toadth; -aois for 4s (final), as gondaois = gontis =nbsp;earlier gontais; c for g, as arcot = argat; cc{h) for g{h),nbsp;as C6ÜCC = ceüg-, p for b, as pa = ba-, p, pli, or xi for b{h),nbsp;as curuut = corbat, feruip = feradb, dipJi = dih; q for cunbsp;as qan = cxian-, occasionally mb for xn, as aixunb = ainm;nbsp;Id for ll, as Maolsecnailld.

He is much given to adding otiose vowels after a final vowel, -ai for -a, -on for -o, etc.: even writing ixnnaix’eaxi fornbsp;ixnndire. He often adds an otiose n, both medial and final,nbsp;writing gavjrb for garb, cauire for coire; one finds such linesnbsp;as la firu Boulc bautar and for xnauco for ingenau, a reliccnbsp;nau raiith-Feniu. When a quatrain ends in the middle of anbsp;line of the manuscript he will often fill up the blank bynbsp;repeating the last letter perhaps a dozen times.

S^. The vellum codex Stowe D. IV. 2 contains, on f. 55 v° and f. 56 r° and v°, seven articles from the Dindshenchas,nbsp;viz.; f, 55, Lusmag (prose and verse). Dun mac Nechtainnbsp;(prose); here at least one folio is missing; f. 56, Ath Lxiaixinbsp;(verse, acephalous), Mag Lena I (pr. and v.). Loch Derg (pr.nbsp;and V.), Sliab Mairge II (pr. and v.), Crechxnael (pr. and v.).nbsp;The last of these is in a different hand from the rest: thenbsp;scribe appends a note at foot of f. 56 v° : / Mainistir Chillinbsp;Corxxiaic daxn. Misi Seaan xnac Aedacain qui scribsit. Cellnbsp;Chormaic (Kilcormick) in Ui Pailge is otherwise known asnbsp;Frankfort in King’s County. The scribe belonged to thenbsp;well-known family of Mac Egans, as to whom see the Introduction to the facsimile of Leabhar Breac, the index tonbsp;Abbott-Gwynn, Catalogue of Irish MSS. in Trinity College,nbsp;Dublin, and O’Donovan, Tribes and Customs of Hy Many,nbsp;p. 168.

The contents of D. IV. 2 are partially described by Meyer (Rev. Celt. VI. 173, seq.). He refers the manuscript to the

GENERAL INTRODUCTION.

late fourteenth century, but it more probably belongs to the fifteenth century.^

The articles seem to be selected quite arbitrarily, as they do not follow the order observed in any copy of thenbsp;Dindshenchas.

The text is nearest to that of B : see critical apparatus to Crechmael, 4, 9, 10, 19 (pt. iv. 22); Loch Derg, 18, 23, 41, 48,nbsp;50, 56, 67, 72, 87, 93,100, 103 (pt. iii. 338, seq.).

II.—First Recension ; (a) The Dindshenchas in the Book

of Leinster.

§ 1. In the present arrangement of the manuscript, which is followed in the published facsimile, the Dindshenchas occupiesnbsp;pp. 151-170 and 191-216; between pp. 170 and 191 twentynbsp;pages of other matter are interpolated. It cannot, however,nbsp;be assumed that this was the original order. The book, as anbsp;whole, had suffered severely before it came into the handsnbsp;of Edward Lhuyd, in or about the year 1700. Much of itsnbsp;original contents was missing, many leaves were crumpled,nbsp;many were detached from their conjugates, some were partiallynbsp;obliterated. Loose leaves had, no doubt, been displaced fromnbsp;their proper positions, and attempts to restore the originalnbsp;order were not wholly successful. We must allow for thenbsp;possibility that not only single leaves, but whole gatherings,nbsp;may have lost their right place. The ten folios (pp. 171-190) which interrupt the sequence of the Dindshenchas formnbsp;a continuous and self-contained section of the MS. Theynbsp;are not connected by contexture of leaves with what precedesnbsp;or wdth what follows, and may have been transferred bodilynbsp;by some rearranger to their present position. It seems morenbsp;likely that such a displacement occurred than that the scribenbsp;turned from his work on the Dindshenchas to write thesenbsp;leaves and then returned to the Dindshenchas.

* In tlie introduction to Ms edition of Merugiid TJilix Meyer dates the codex 1300 a.d., relying on an entry on fo. 1. But the figuresnbsp;“MCCC” are not, as he asserts, part of the Latin note in the handnbsp;of the text: they are entered separately in a different and, apparently,nbsp;a later hand, and may safely be ignored.

GENERAL INTRODUCTION.

Further, it has long been recognised that the order of the Dindshenchas itself is wrong at two points. The first halfnbsp;of the poem on Brug na Bóinde is on p. 164, the latter halfnbsp;on p. 211; and the poem on Berha, which begins on p. 216,nbsp;is completed on p. 191. A closer study of the contexture ofnbsp;leaves, and of the script, will brinff out some other pointsnbsp;deserving of consideration.

Pp. 151-154 are conjugates of 143-146; their position is therefore secure.

Pp. 155-158 (two detached leaves) follow these without break of continuity between page and page. P. 158 endsnbsp;with a blank space. So far all is in verse.

Pp. 159-160 begin the prose. This leaf is conjugate with 165-166, but 165' begins with an acephalous article, so thatnbsp;we must mark a gap between 160 and 165.

(Pp. 161-164 are out of order: see below.)

P. 167 continues 166; 167-168 is a detached leaf and 168 ends with a completed article, so that there may possibly benbsp;a gap between this leaf and the next, but I see no reason tonbsp;assume that such a gap exists.

Pp. 169-170, another detached leaf (here follow the ten interpolated leaves, pp. 171-190).

Pp. 215-216, a loose leaf, must, as we have seen, be placed before 191. As 215 begins in the middle of a prose articlenbsp;{Carmun),^’^ we must mark another gap before this leaf.

Pp. 191-202 consist of three pairs of conjugates (191-2 = 201-2, 193-4 = 199-200, 195-6 = 197-8).

Pp. 203-6, two detached leaves.

*a.The article begins witt the paragraph Sengarrmn tanic i ndegaid and above it is written vel ita. In MSS. of the Second Recension thisnbsp;parag-raph is preceded by four others (as in Rev. Celt. xv. 311) foundednbsp;on 11. 17-76 of the poem on Carman, and vel ita is written over par. 5,nbsp;as in L. I assume that a leaf is lost before p. 215 which containednbsp;these paragraphs. Thurneysen (Heldensage) does not admit the loss ofnbsp;a leaf here. If he is right, we must suppose that paragraphs 1-4 of thenbsp;Second Recension were added by the Reviser, and must explain the velnbsp;ita as intended by the scribe to distinguish the prose from the metricalnbsp;version which follows. This seems unlikely, yet we have a somewhatnbsp;similar case at p. 193a where the dindshenchas ot Fornocht is given bothnbsp;in prose and in verse, and over the prose is written vel aliter, though thenbsp;prose precedes the verse.

GENERAL INTRODUCTION.

Pp. 207-8 are conjugate with 211-2.

Pp. 209-10, a loose leaf, ending with a blank space.

Prom 215 to the end of 210 there is no break of connexion between page and page.

Pp. 161-4 form two detached leaves, but their contents are continuous, and we have seen that 164 must immediatelynbsp;precede 211: these four pages must therefore be placednbsp;between 210 and 211.

Pp. 213-4, la loose leaf, but 213 continues 212. The Dindshenchas ends with 214.

Thus we arrive at the following order:—

I. nbsp;nbsp;nbsp;151-158. All verse.

II. nbsp;nbsp;nbsp;159-160, •— , 165-170. All prose except for the poemsnbsp;on Lagin, and a few scattered stanzas: see p. 16, below.

III. nbsp;nbsp;nbsp;— , 215-6, 191-210, — (?), 161-4, 211-214. Mainlynbsp;verse, with a few prose articles: see pp. 17-19, below. Thisnbsp;section includes several poems, especially on pp. 205-8, whichnbsp;do not properly belong to the Dindshenchas.

It is possible that II and III originally stood in reverse order, the metrical part being completed (except for Lagin)nbsp;before the prose was begun. There is nothing an the contexture of the leaves to disprove this supposition.

§ 2. The handwriting offers two points of interest.

(I). The script of pp. 161-4 is remarkably line and careful: the letters a- and d are particularly noticeable, andnbsp;the sharp angles are finished with more care than in anynbsp;other part of the Dindshenchas. There are scarcely any othernbsp;leaves in the whole manuscript which equal these two innbsp;penmanship; accordingly one of these leaves was chosen fornbsp;photographic reproduction in the facsimile, where it gives anbsp;somewhat too flattering impression of the volume as a -whole.nbsp;Also the vellum seemsi to me to be somewhat smoother andnbsp;finer than that employed for the rest of the Dindshenchas.nbsp;When one compares these two leaves with what precedes andnbsp;follows, one is at first disposed to think that they are the worknbsp;of a different scribe; but a careful examination of the lowernbsp;part of 161a, where the lines are crowded and the handnbsp;deteriorates, has convinced me that these pages are the work

GENERAL INTRODUCTION.

of the same scnibe (“A”) who wrote the rest of the Dindshenchas, both prose and Verse, with the exception ofnbsp;the columns which I have next to mention.

(2). The second column of 208, with all that follows in the present order of pagination down to the end of 214, isnbsp;in a hand (“B”) markedly inferior to that of the main scribe,nbsp;even when his work is most hurried and crowded. This handnbsp;is shaky and uneven, and offers a striking contrast to thenbsp;beautiful clearness and regularity of the script in 161-4. Thenbsp;difference between the columns written in this hand and thosenbsp;that precede cannot be attributed to any roughness of thenbsp;vellum. The contrast between the fine penmanship of thenbsp;first half of the poem on Brug na Bóinde on 164 and thenbsp;conclusion in B’s hand on 211 is particularlj^ striking.

I believe that A broke off at the end of the first column of 208, and that B continued the task of transcription to thenbsp;end of 210. At this point A inserted the two leaves 161-164,nbsp;which he had for_ some reason written with more care andnbsp;probably at a different time from the rest. B then resumednbsp;work on the pair of vellum sheets which he had been using,nbsp;completing a poem left unfinished by A on p. 164, andnbsp;carrying on to the end of the verse Dindshenchas. If thenbsp;prose originally followed the verse, as has been suggestednbsp;above, A must have relieved B after p. 214.^

§ 3. We have now to see what further light can be obtained from an examination of the contents of the various sectionsnbsp;into which the L-Dindshenchas falls. We shall have tonbsp;distinguish, not as at first three sections, but five or perhapsnbsp;six. I proceed to list the whole contents of the L-Dindshenchas,nbsp;verse and prose, noting the province to which each place

^ On making a fresh examination of the whole manuscript I find similar apparent changes of hand on p. 21h (the poem A Emain, afternbsp;the first three words), p. 22 (the poem Ugaine uallach, all but the firstnbsp;three words), pp. 27-30, 37, 59-42, 50 {Kaeo sunt nomina uirorumnbsp;componentium lapides), and 51 6 38 to 52 6 25. All these passages seemnbsp;to be in the hand of B: but I now think it possible that B is merely A,nbsp;writing under specially bad conditions. If so, there is a surprisingnbsp;difference between his best work and his worst. The poem Sian seissnbsp;on 496 and the tale Luid Feidilmid Bechtaid on 51a seem to be writtennbsp;by another hand, distinct from A and B.

GENERAL INTRODUCTION.

belongs. Many of them have not yet been identified, but we may in uncertain eases be guided by the position they occupynbsp;in the Second Recension, which is arranged, as we shall see,nbsp;though imperfectly, in geographical order. We have alsonbsp;evidence as to the provinces to which unidentified placesnbsp;belong in 0’Duinn’s poem (see p. 77, below). But 0’Duinnnbsp;is by no means a reliable guide: he is certainly mistaken innbsp;placing Mag Muiresce (62) in Meath and Benn Boguine (40)nbsp;in Connaught.

Loch Derg and Ath Liiain are here treated as belonging to Connaught, because they are so regarded in the Secondnbsp;Recension.

TODD LECTURE SERIES, VOL. XII.

§ 4. It is evident that there is no continuous geographical arrangement throughout the series as a whole. But it is alsonbsp;clear that there are signs of attempts at such an arrangement. Thus in I, 5 places in Munster (7-11) are followednbsp;by 8 in Connaught (12-19). In this section there arenbsp;altogether 9 Connaught names, 5 from Munster and 2 fromnbsp;each of the other provinces. In III (a) (omitting the pro.senbsp;of II for the moment) Leinster predominates. The beginningnbsp;of this section is lost: possibly it contained a Meath series.nbsp;As it stands, it begins with 35 places, of which 3 are doubtfulnbsp;(89, 101-2), and the rest (counting Bound: the river risesnbsp;in Kildare) are in Leinster. Of the next 8 places (122-129),nbsp;7 are in Munster, 1 in Leinster (124); then come 6 scatterednbsp;places; then come 5 in Leinster, then 8 scattered over thenbsp;five provinces. Here the scribe interposes a number of poems

GENERAL INTRODUCTION.

which do not belong to the Dindshenchas, and then adds

3 nbsp;nbsp;nbsp;places which seem to close a section of the work. In thisnbsp;section there are 42 places from Leinster, 11 from Munster,

4 nbsp;nbsp;nbsp;from Connaught, 3 from Meath, 2 from Ulster, and 5 arenbsp;uncertain.

In Section III (b) (pi). 161-4) there is no trace of geographical order, but when we pass to III (c) we find anbsp;marked preponderance of Connaught names; 11 out of 16nbsp;belong to that province, not counting Loch Erne and Essnbsp;Ruaid, which are on the confines of Connaught and Ulster.

§ 5. The facts stated above (§ 4) seem to indicate that the verse-Dindshenchas in L is derived from several separatenbsp;documents, among which one (Ilia) was mainly concernednbsp;with Leinster, and another (IIIc) mainly with Connaught,nbsp;If the whole collection had been made by a compiler whonbsp;intended to put together the Dindshenchas of all Ireland,nbsp;without having any previous partial attempts to work upon,nbsp;he would hardly have begun in the abrupt and random fashionnbsp;in which the scribe of L attacks his task. There has been anbsp;process of accretion comparable to that by which Irishnbsp;dictionaries have been formed. One compiler found and putnbsp;together glossaries of separate texts some of which followednbsp;the order in which excerpts were made, others the order ofnbsp;the letters of the alphabet; then a later compiler, likenbsp;0’Davoren or 0’Clery, attempted to digest the whole on thenbsp;alphabetical plan; the final result being a mixture of systemnbsp;and haphazard. It can hardly have been the scribe of L whonbsp;thus formed the Dindshenchas Érenn from smaller units. Atnbsp;first sight one is tempted to give him the credit and tonbsp;suppose that the blanks which he has left at the foot ofnbsp;p. 158 and p. 210 mark the end of two of the documentsnbsp;which he was utilising. But this supposition cannot well benbsp;reconciled with the conclusions arrived at below (p. 67) asnbsp;to the relations between the Reviser’s model and the text ofnbsp;L. It will there be shown that the Reviser had before himnbsp;a text in many points different from and superior to L : henbsp;cannot have worked directly from L. If the scribe of Lnbsp;collected for the first time five or six different documents, wenbsp;should have to suppose that the Reviser had access to all of

GENERAL INTRODUCTION.

these and made (on the whole) better use of them. This is a very improbable supposition. We must then assume thatnbsp;there existed a collection prior to L, from which L and thenbsp;revised text of the Second Recension are independentlynbsp;derived. There is definite proof of this. In the poem onnbsp;Inber Ailbine there is a hiatus of four lines (the latter halfnbsp;of one stanza and the first half of the next) which is commonnbsp;to L and to the whole Second Recension. It was thereforenbsp;already in the MS. to which both the scribe of L and thenbsp;Reviser had access.

As to the blanks in L noted above, the poem on Maistiu, which is the last entry on p. 158, is unfinished, and the blanknbsp;which follows may, as Thurneysen suggests, have been leftnbsp;by the scribe with the intention of completing the poem fromnbsp;another copy. The blank at the foot of p. 210b remainsnbsp;unexplained; but it may be noted that at other points thenbsp;scribe left blanks for no apparent reason. On p. 163 therenbsp;was a vacant space after the poem on Diin Mdse, which wasnbsp;filled by a later hand with the paragraphs on Duma Oenanbsp;and other places. So too at the foot of p. 201b the entrynbsp;about Cnoc Rafann is a later addition on a space originallynbsp;left blank. There is also a smaller vacant space at the footnbsp;of p. 168a.

§ 6. We have now to consider the prose section. Like the verse it shows a half-hearted attempt at a geographicalnbsp;arrangement. Beginning with Temair—a choice whichnbsp;suggests that the compiler had in view a collection representative of all Ireland—it proceeds with a series of 13 placesnbsp;in Leinster. Then after a gap in the manuscript we havenbsp;5 places in Connaught (35-39), and from 53-63 another seriesnbsp;of 11 Connaught names (but 60 is a doubtful case). Promnbsp;65-78, all are in Munster, with the probable exceptions ofnbsp;69 and 76.

It can hardly be doubted that the prose Dindshenchas has been put together largely by making abstracts of the corresponding poems. Some of the poems appear from theirnbsp;language to be older than the prose (e.g., those on Ath Macnbsp;Find, Ailech I, Mimman Tige Srafdin, Benn Étair I). Somenbsp;are ascribed to authors of the early eleventh or even of the

GENERAL INTRODUCTION.

tenth century {e.g., Achall, to Cinaeth ua Hartacain, t974), and we have no reason to put the composition of the prosenbsp;so early. In several eases the prose compiler quotes the firstnbsp;line or detached stanzas of a poem which belongs to the versenbsp;Dindshenchas, and which is evidently his source : the instancesnbsp;are Loch Garman, SUab Bladnia, Mag Raigne (all on p. 159),nbsp;Dun Mdse, and Faffanji (p. 160), Nenithenn (p. 165), Echtganbsp;(p. 170), besides Rath Chndmrossa and Cend Finichair (p.nbsp;200), which are included in one of the verse-sections. Alsonbsp;the prose of Loch Riach ends with reference to a (lost) poemnbsp;beginning Rochuala cocad n-amra. We may add two casesnbsp;where the prose refers to a poem which no longer exists innbsp;L, but is found in the Second Recension : these are Fichnbsp;Buana (p. 166) and Findglais (p. 169). In these instances,,nbsp;however, the prose-writer seems to draw directly from thenbsp;well-known tales Fled Bricrenn and Aided Chonroi rathernbsp;than from the poems which he mentions. Speaking generally,nbsp;when we have both the prose and the verse legends of anynbsp;place in L for comparison, we find that the prose is usuallynbsp;no more than a brief extract from the poem, while in mostnbsp;cases the poem contains much that is not to be found in thenbsp;prose. It is true that the prose often gives names andnbsp;incidents not derived from the verse: such additions are nonbsp;doubt drawn from the compiler’s own knowledge of thenbsp;common stock of legendary lore.

§ 7. There are 64^“ poems in the verse of the L-Dindshenchas for which there is no prose equivalent, and there are 30 prosesnbsp;for which there is no corresponding poem. As both prosenbsp;and verse are defective, owing to losses of folios, we maynbsp;presume that many of the missing equivalents would benbsp;supplied if the manuscript were still complete. But it wouldnbsp;be a mistake to assume that the original collection possessednbsp;both verse and prose for every place on the list. Comparisonnbsp;with the Second Recension shows that 25 out of the 30 prosesnbsp;just mentioned are there provided with metrical versions.nbsp;Most of these, however, as Thurneysen rightly insists, arenbsp;clearly later work, and are probably due to the Reviser. The

”a Not reckoning the poems referred to below, p. 25, nor two or three which appear in L, but not as part of the Dindshenchas.

GENERAL INTRODUCTION.

relation between these poems and the prose is the converse of that which we find in L. For the most part they arenbsp;evidently based on the prose, adding no more to its datanbsp;than some metrical padding. We must infer that thenbsp;Reviser in these cases found no poem in the older recensionnbsp;and himself supplied a metrical version.^ The poems in thenbsp;original Second Recension (ending with Bile Tortan, no. 167nbsp;of the list on p. 48, below) which are not also found in L arenbsp;61 in number. For many of these no equivalent exists in L’snbsp;prose in its present state. But if w’e assume that the earlynbsp;prose Dindshenchas, when complete, included those legendsnbsp;which are found in the Bd.-Bd. recension, though no longernbsp;in L, then we have prose equivalents for all but 13 of thesenbsp;61 poems. The 13 which remain unaccounted for are:nbsp;Temair III, Brug na Bóinde I, Inher nAilhine, Dun Crini-thainn, Temair Luuchra, Mag nAi, Duma Selga, Athnbsp;nGrenclia, Carraic Letlideirg, LecJit Oenfhir Aife, Loch Ddnbsp;Gabar, Mag Léna, and Bile Tortan : and in the first twonbsp;installces, L has a different poem on the same subject.

§ 8. As already stated, L has 64 poems without an equivalent in its prose.^

For 13 of these prose versions are supplied by the Bd.-Ed. recension: these proses therefore (or most of them) probablynbsp;existed also in L in its complete condition. For most of thenbsp;remaining 51 poems the Reviser supplied a prose version, justnbsp;as he supplied metrical versions where these were lackingnbsp;in L; or else he adapted versions then existing in the Firstnbsp;Recension, but now lost.

’ The four proses which want a verse equivalent in L, and were omitted by the Reviser in the original draft of the Second Recension,nbsp;are Srub Bo, Mag Tarbga, Sêig Mossad, Loch nOirhsen.

The case of Ceilbe is peculiar. The prose in L is introduced by the words ut ante, uel aliter, as if the scribe had already copied an entrynbsp;about this place. Was this a poem, which stood in the leaf that hasnbsp;been lost before p. 215? If so, it was omitted by the Reviser, whonbsp;only copied the prose legend. In two late MSS. it was fitted with anbsp;poem by an author of the sixteenth century.

In two more instances places which have both verse and prose legends in L are omitted from all MSS. of the Second Recension.nbsp;These are Cenn FinicJiair and ’Lurnman Tige Srafdin.

' Note that some of the proses accompany the poems (see list on pp. 17-19). In no case are these repeated in L’s prose-collectionnbsp;(Thurneysen’s “Ba”).

GENERAL INTRODUCTION.

Some of the L-poems were, however, passed over by the Reviser. Thurneysen supposes that these did not belong tonbsp;the original Dindshenchas, but were added by the scribe ofnbsp;L (Heldensage, 38-9 and 43). It seems, however, quite asnbsp;likely that the Reviser excluded them, at least in some eases,nbsp;because they did not conform to his stricter conception ofnbsp;the term dindshenchas.

On the same ground a number of poems are omitted from the present edition which the scribe of L has inserted amongnbsp;the true Dindshenchas poems. Most of these occur betweennbsp;p. 204 and p. 208 of L. Thurneysen treats them as part ofnbsp;the collection, but they cannot be brought within any definitionnbsp;of the term dindshenchas, however loosely interpreted. Inbsp;regret, however, that the poem on the legend of Bóand (pp.nbsp;208-9) has also been omitted. It ought to have been includednbsp;in my second volume, after Bóand II.

III. First Recension : (b) Bd.-Ed.

§ 1. Two MSS., Rawl. B. 506 (“Bd.”) and Kilbride XVI (“Ed.”) contain a recension of the prose Dindshenchas,nbsp;differing both from that of L and from the Second Recensionnbsp;in contents, in arrangement, and in text. Both arenbsp;fragmentary; they supplement each others’ deficiencies to anbsp;considerable extent, but the full list of the original contentsnbsp;remains a matter of conjecture. The contents of Bd. andnbsp;those articles in Ed. which are not found in Bd. have beennbsp;edited by Stokes, Folk-Lore, III (1892), pp. 467 seq., and IVnbsp;(1893), pp. 471 seq.® A description of the state of bothnbsp;manuscripts, with conjectures as to their original contents,nbsp;will be found in Stokes, ut sup., and in Thurneysen’snbsp;Heldensage, pp. 39-43. In an article contributed to Ériu X,nbsp;68-74, it has been shown that in the Dindshenchas in thenbsp;Book of Ui Maine (M) there are interpolated some 20 articles,nbsp;the text of which approximates to that of Bd.-Ed. Fornbsp;purposes of comparison with the other recensions I append anbsp;table of the present contents of both manuscripts, and shownbsp;the correspondences with the M-interpolation, with L (prosenbsp;and verse), and with the Second Recension.

' He gave the classing of Ed. wrongly as “Kilbride XVI,” a mistake which is repeated in the present edition, pt. iii. p. viii, pt. iv. p. ix.

TODD LECTÜKE SERIES, VOL. XII. nbsp;nbsp;nbsp;D

26 nbsp;nbsp;nbsp;GENERAL INTRODUCTION.

§ 2. Contents of Bd. and Ed.

GENERAL INTRODUCTION.

GENEEAL INTRODUCTION.

It will be seen that the surviving part which is common to Bd. and Ed. is arranged in both in the same order, exceptnbsp;that Tmg Inbir is placed in Bd. before Cleitech, in Ed.nbsp;before Benn Boguine. Further, the text of the two manuscripts agrees so closely that it is clear that we have to donbsp;with two fragmentary copies of one document This assumptionnbsp;is confirmed by the fact that the M-interpolation has sixnbsp;items which are in Bd. but missing in Ed., and seven whichnbsp;are in Ed. but missing from Bd.

When the Bd.-Ed. text is compared with the prose of L and of the Second Recension, in the articles which are commonnbsp;to all three, it is found as a rule to be much closer to thenbsp;former than to the latter. In fact, in a large proportion ofnbsp;cases Bd.-Ed. and L are practically identical, apart fromnbsp;minor additions and omissions: whereas Rev. is usuallynbsp;differentiated by adding (or less frequently, by omitting)nbsp;aliter versions of the legend, or by retelling in different wordsnbsp;the versions which it has in common with L and Bd.-Ed.nbsp;On the other hand there are many instances of agreementnbsp;between Bd.-Ed. and Rev. against L. In an article contributednbsp;to Hermathena (1932) I studied the relations of the textnbsp;common to Bd., Ed., and the M-interpolation with the prosenbsp;of L on the one hand and with that of Rev. on the other.nbsp;The conclusions reached were—

(i) nbsp;nbsp;nbsp;that Bd.-Ed.-M derive from a text (j8) which was itselfnbsp;not a complete copy of the Dindshenchas, but a selectionnbsp;based on a manuscript of the First Recension (a), othernbsp;than h;

(ii) nbsp;nbsp;nbsp;that /8, although an off-shoot of the First Recension,nbsp;represents a later stage of the tradition than L, containingnbsp;certain additions and developments, some of which werenbsp;probably already present in a;

(iii) nbsp;nbsp;nbsp;that the Reviser’s work was based on a, though henbsp;used it rather as a quarry for material than as a model.

(Note that the symbol (3 was in the Hermathena article employed otherwise than in the present volume, in which itnbsp;is used to denote the common ancestor of B and R.)

GENERAL INTRODUCTION.

IV.—Second Recension.

§ 1. The compiler of the Second Recension had a stricter notion of the meaning of the term dindsJienchas than thenbsp;first compiler, and his work is throughout more systematic.nbsp;He brings together the prose and metrical versions of eachnbsp;legend, usually separated in the older arrangement, alwaysnbsp;putting the prose first, and introducing it with the unvaryingnbsp;formula : ‘X unde nominatur? rd insa.’ Poems which didnbsp;not conform to his definition he seems to have omitted onnbsp;principle, and this is no doubt the reason of the exclusion ofnbsp;several legends included in the First Recension. When eithernbsp;a pros© or a metrical version was wanting, he seems to havenbsp;supplied the deficiency from his own resources' (see II, §§ 7, 8).nbsp;Further, it is clear that he intended from the first to follownbsp;a. geographical arrangement (and it may be noted that thenbsp;additions made in later copies of the Second Recension follownbsp;tlie same principle, with few exceptions). MTien he foundnbsp;clear traces of an attempt at geographical sequence in thenbsp;verse of the earlier recension he has preserved its order, as thenbsp;following examples will show.

Belach Durgein Bairenn Chermainnbsp;Dublindnbsp;Fornocht

GENERAL INTRODUCTION.

2. L-Eeeension, 110^122—contd. nbsp;nbsp;nbsp;Second Recension, 28-48—contd.

Thus the Keviser has preserved all the longer sequences in the older verse recension,® interpolating, however, a fewnbsp;places according to his own ideas of geographical fitness.

This principle has determined the arrangement of the first 114 numbers of the first draft (BE), which are distributednbsp;according to provinces as follows. First come 7 places innbsp;Meath, then 32 (counting Bórnid) in Leinster. Next we havenbsp;20 places (nos. 44-64), all in Munster, except 48 {Mag Raigne)nbsp;and 55 {Cenn Currig), which are in Leinster, and 60 {Srubnbsp;Brain), as to which see Heldensage, p. 492. Then come 31nbsp;places (nos. 65-95) in Connaught. After these no. 96 {Bruimnbsp;Criaich in Meath) is inserted, out of geographical order.nbsp;Then 16 places (nos. 97-113) mostly in Ulster, but 100 {Athnbsp;nGrencha) is probably in Meath and one or two more arenbsp;doubtful. Here the compiler’s procedure becomes morénbsp;difficult to follow. As to the 32 places, nos. 114 to 145 (at

” At one point, from 132 to 137 (Mag Findahrach-Inier Bicne), tlie Reviser has followed the order of L’s prose.

GENERAL INTRODUCTION.

which point the Second Recension seems originally to have ended) it is to be remarked, first, that only one (114) occursnbsp;in L; secondly, that most of the places, so far as they havenbsp;been identified, seem to belong to Meath, but that there is annbsp;admixture of names from all the other provinces. We maynbsp;assign to Meath with more or less certainty the following :nbsp;Tailtiu, Odha, Loch Bd Gabar, Ben-n Chodail (?), Tlachtga,nbsp;Mag Breg, Cleitecli, Gerna, Irarus, Mag Findabrach, Gdirech,nbsp;Trdig Tuirhe, Bri Léith, Tetliba, Loch nAindind, Druimnbsp;Simnaig, Bun Mac Nechtain (?), Bile Tortan: in all, 18nbsp;places. To Ulster belong perhaps 6 places, Sliab Fuait, Sliabnbsp;Callann, Inber Cickmaine (?), Fich Buana (?), Lia Linga-dain (?), Inber Bicne, of wRich only 3 are certain. Tonbsp;Connaught we may assign at most 3 places, 3Ióin Tire Ndir,nbsp;Odras, Cloenloch. Only three, Sruthar Matlm, Lidbnech andnbsp;Loch Seta, probably belong to Munster, and two, Lusmag (1)nbsp;and Mag Léna, to Leinster.

There is thus a marked predominance of Meath names. Why did the Reviser thus turn back to his starting point?nbsp;A reason suggests itself when we refer to what is said (II § 7)nbsp;of the sources of his compilation. It is there assumed thatnbsp;for the most part he either (a) took the poems which he foundnbsp;in the older recension, or (b) composed fresh poems bynbsp;versifying the prose of the older recension, when this contained no corresponding poem. But he made little use ofnbsp;the second method until he had advanced some way in hisnbsp;task. Out of the first 91 places celebrated in the Secondnbsp;Recension, in its original form and omitting later additionsnbsp;(and also omitting Ceilbe, which had at first no metricalnbsp;version), there are only 12 not represented in the verse of L,nbsp;namely, Inber Ailbine, Bun Crimtiminn, Mag Mugna, Lochnbsp;Bachaech, Tonn Chlidna, Crotta CUach, Temair Luachra,nbsp;Findglais, Moenmag, Mag nAi, Biima Selga, and Loch Coni

' And some of these probably had their places in the folios which I assume to be lost before p. 215 of L. On the other hand, the poemnbsp;on Almu in the Second Recension is quite different from that in L,nbsp;and the Second Recension adds alternative poems on several placesnbsp;besides those in L

GENERAL INTRODUCTION. nbsp;nbsp;nbsp;33

It is only after No. 98 {Benn Boguine) that the Reviser begins to add numerous poems versifying L’s prose legendsnbsp;or drawing from his own resources. These are at firstnbsp;concerned with places in Ulster, which province is, as we havenbsp;seen, poorly represented in L. Afterwards (from 117 onward)nbsp;he seems to have turned back to the remaining prose legendsnbsp;for which the earlier recension had no verse equivalent, andnbsp;to have given them a metrical form. The greater number ofnbsp;these were, as has been remarked, attached to places in thenbsp;province of Meath.

§ 2. The table which follows shows the contents of all MSS. belonging to the Second Recension, with the exceptionnbsp;of such as are mere copies of other extant MSS., and also ofnbsp;V and X, which are late and valueless. MSS. which containnbsp;short excerpts from the Dindshenchas, without regard tonbsp;order, are not included (see p. 4). The columns are spacednbsp;so as to show correspondences and differences of arrangement.nbsp;The numbering in the first column relates to the completenbsp;list: that in the second column is confined to the originalnbsp;contents of BR, which I regard as the first form of thisnbsp;Recension.

[Table.

TODD liECTUKE SERIES, VOL. XII.

GENEEAL INTRODUCTION.

K 2

GENERAL INTRODUCTION.

M

Bri Léith Tethbanbsp;Loch Aindindnbsp;Lusmagnbsp;Benn Chodailnbsp;I. Cichmainenbsp;Loch Gabar

Loch Ri II (see iii. .560)nbsp;Dubtbir

GENERAL INTRODUCTION.

H

Mag Finnabrach Druim Dairbrech

GENERAL INTRODUCTION.

GENERAL INTRODUCTION.

GENERAL INTRODUCTION.

GENEEAL INTKODUCTION.

GENERAL INTRODUCTION.

GENERAL INTRODUCTION.